338 
PIJOFESSOK 0. MASSON ON IONIC VELOCITIES. 
the bath temperature, which sliould keep constant. With constant E.M.F. the 
current and the velocities steadily diminish in a manner determined by the nature 
and strength of the salt-jelly, hut they remain in direct proportion to one another. 
The exjDeriment is terminated when the two boundaries meet, which they do at a 
})oint that can always be predicted with very considerable accuracy from the first 
readings. 
The essential pi'inciple of the method may now be explained. The visible moving 
Ijoundai'ies mark respectively, not only the rates of advance of the foremost Cu and 
Cr 04 , ljut also those of the rearmost K and the rearmost Cl. These are themselves 
invisible, ljut the immediately following coloui’ed ions may be taken as their indicators. 
Now the intermediate colourless part of the jelly is at the start of uniform composition 
and remains so throughout the experiment, however much its length may he curtailed 
Ijy the progress of the blue and the yellow; so that (to use the same symbols as 
Ijefore) n and x have constant values from start to finish and in all parts of the 
colourless jelly, while tt, which diminishes as the experiment proceeds, has yet always 
the same value there, whether the part near the blue boundary or the part near the 
yellow boundary be considered. Therefore the i-earmost K at the one end and the 
rearmost Cl at the other are comparable in all respects ; and a comparison of their 
working velocities U and V, made visible by the indicators, gives at once the ratio u/v 
for the particular- concentration employed. Obviously also the result may be put in 
the form of Hittorf’s transport number ^ (or p), and may be compared with the 
values obtained by the indirect Hittorfian method. Of course, the observed U is 
also that of the copper ions and the observed V that of the CrO^ ions ; hut the 
experiment affords no indication of the corresponding values of tt and ic, which are 
certainly very different in the two cases. Hence there is nothing gamed by regarding 
the observations from this point of view, as has already been pointed out in connection 
with Lodge’s experiments. 
While the first result is the determination of u/v for the original salt, a second is 
the testing of the general equation C = A (U -}- V), or, if its truth he assumed, of 
the efficacy of the method itself For each quantity is independently determined, 
and, since all may he expressed in the same (C.G.S.) units, the value 1 should he 
obtained by dividing the left-hand side by the right. 
The experiment, as carried out, also affords data for the determination of the 
working velocities per unit potential slope, viz., of U/tt = xu and V/tt = xi\ For the 
total resistance for each position of the boundaries is given by the readings of voltage 
and current; and, as the increase of total resistance is directly proportional to the 
diminution of the length of the colourless jelly, and as a constant correction can be 
introduced for the resistance of the galvanometer, and approximately also for that of 
the solutions between the electrodes and the ends of the tubes, the resistance of the 
